bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Early candidate urine biomarkers for detecting Alzheimer’s disease before beta amyloid plaque deposition in an APP (swe)/PSEN1dE9 transgenic mouse model

Fanshuang Zhang1,2*, Jing Wei3*, Xundou Li1*, Chao Ma4#, and Youhe Gao3#

1Department of Pathophysiology, Institute of Basic Medical Sciences Chinese Academy of Medical

Sciences, School of Basic Medicine Peking Union Medical College, Beijing, 100005, China

2Department of Pathology, National Cancer Center/Cancer Hospital, Chinese Academy of Medical

Sciences and Peking Union Medical College, Beijing, 100021, China

3Department of Biochemistry and Molecular Biology, Beijing Normal University, Gene Engineering

Drug and Biotechnology Beijing Key Laboratory, Beijing, 100875, China

4Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences, Department of Human

Anatomy, Histology and Embryology, Neuroscience Center; Joint Laboratory of Anesthesia and Pain,

School of Basic Medicine, Peking Union Medical College, Beijing, 100005, China

∗These authors contributed equally to this work.

#Corresponding author: Youhe Gao & Chao Ma

Email: [email protected]; [email protected]

Phone: 86-10-5880-4382; Fax: 86-10-6521-2284

Abstract

Alzheimer’s disease (AD) is an incurable age-associated neurodegenerative disorder that is

characterized by irreversible progressive cognitive deficits and extensive brain damage. The

identification of candidate biomarkers before beta amyloid plaque deposition occurs is therefore of

great importance for the early intervention of AD. Urine, which is not regulated by homeostatic

mechanisms, theoretically accumulates changes associated with AD earlier than cerebrospinal fluid and

blood. In this study, an APP (swe)/PSEN1dE9 transgenic mouse model was used to identify candidate

biomarkers for early AD. Urine samples were collected from 4-, 6-, and 8-month-old transgenic mouse

models, and the urinary proteomes were profiled using liquid chromatography coupled with tandem

mass spectrometry (LC-MS/MS). The levels of 33 proteins differed significantly between wild-type

and 4-month-old mice, which had not started to deposit beta amyloid plaque. Among these proteins, 16

have been associated with the mechanisms of AD, while 9 have been suggested as AD biomarkers. Our bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

results indicated that urine proteins enable detecting AD before beta amyloid plaque deposition, which

may present an opportunity for intervention.

Key words: Alzheimer’s disease (AD), urine proteome, early diagnosis, APP (swe)/PSEN1dE9

Introduction

Alzheimer’s disease (AD) is a chronic age-associated neurodegenerative disorder associated with

cognitive impairment and progressive dementia[1]. As the pathological process leading to AD begins

decades before clinical symptoms, finding early clues in the early stages of AD, especially before beta

amyloid plaque deposition, is urgent in today’s AD research. Most candidate AD biomarkers came

from cerebrospinal fluid (CSF) and blood studies; for example, tau levels increase in the CSF of most

AD patients, and stable miRNAs in human serum are potentially valuable novel biomarkers for the

diagnosis of AD. However, beta amyloid deposition can appear in AD patients, even in the stage 1 of

preclinical AD, and can be detected by positron emission tomography (PET) amyloid imaging or CSF

amyloid beta levels[1].

Biomarkers are measurable changes associated with disease. Unlike blood, which is stable because of

homeostasis mechanisms, urine can accumulate many kinds of changes; some of these changes are

associated with disease and will become biomarkers[2,3]. In addition, urine is a resource for biomarker

discovery that easily and non-invasively collected from AD patients. As a sensitive biomarker source,

urine therefore might reflect pathological changes, especially in the early stages of disease[4]. However,

whether time-course analyses of urine proteins at different AD phases can reveal sensitive biomarker

clues before beta amyloid deposition is unclear, as urinary proteins are easily affected by complicated

factors such as medicine and diet, especially in human samples. Using animal models is therefore the

most efficient way to establish an association with disease, as animal models can reduce the influence

of genetic and environmental factors on the urine proteome to the greatest degree[2]. In addition,

animal models can represent the early stages of disease without clinical symptoms as it is impossible to

clinically collect such samples[5]. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Several studies have applied proteomic[6] and metabolomic[7,8] analyses to characterize candidate

biomarkers in mouse models. However, these studies used either plasma or brain tissues, and analyses

of these materials were not sensitive enough for the early diagnosis of AD (before beta amyloid plaque

deposition). Other studies have illustrated urine markers using samples from AD patients[9]. However,

all these studies were conducted after beta amyloid deposition had appeared in brain tissue and

cognitive impairment or movement disorder symptoms had occurred. CSF has a greater homeostatic

priority than blood and changes less than blood. Urine may thus be a better choice for the detection of

earlier biomarkers, even for brain diseases[10]. In addition, the limited studies about urine-based

biomarkers of brain diseases suggest that urine is an ideal potential source for biomarkers in brain

diseases[11]. For example, urine biomarkers in an astrocytoma rat model can reflect early astrocytoma

changes before Magnetic Resonance Imaging (MRI) and thus gives potential clues for the early clinical

diagnosis of astrocytoma patients[12].

Amyloid precursor protein (APP) (swe)/PSEN1dE9 transgenic mice, overexpressing mutant APP and

PS1 (APP/PS1), have been widely used as a model of AD to elucidate the pathogenic processes of the

disease and to investigate candidate biomarkers[13]. Deposition of beta amyloid in the hippocampus of

6-month-old mice occasionally occurred[14,15], and a mass of beta amyloid had been deposited in the

hippocampus of 8-month-old mice [16,17]. Compared with wild-type mice, the 8-month-old transgenic

mice were mainly impaired in the foreground fear conditioning test and in plasticity to acquire

strategies to swim toward a cued escape platform[18]. The 12-month-old transgenic mice showed

spatial learning deficits as well as long-term contextual memory deficits and large deposits of

aggregated insoluble beta amyloid[19]. The cognitive abnormality of APP (swe)/PSEN1dE9 transgenic

mice appeared at 4 months, when beta amyloid plaques had not yet deposited[20]. Potential candidate

urinary biomarkers that appear before brain pathology can thus be identified by using 4-month-old APP

(swe)/PSEN1dE9 transgenic mice. Figure 1 presents the characteristics of behavioral and pathological

profiles of different stages of the APP (swe)/PSEN1dE9 transgenic mouse model.

bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Figure 1. Characteristics of behavioral and pathological profiles of different stages of APP (swe)/PSEN1dE9 transgenic mice

In this study, the APP (swe)/PSEN1dE9 transgenic mouse model was used to identify candidate early

AD biomarkers. To identify changes in the urinary proteome during AD development, we collected and

analyzed urine samples of APP (swe)/PSEN1dE9 transgenic mice at three growth stages (4-, 6- and

8-month-old). The workflow of the proteomic analysis in this study is shown in Figure 2. By using

label-free and TMT-labeling proteomics analysis, we analyzed differential urine biomarkers from

different stages of AD.

bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Figure 2. Workflow of urinary proteomics in APP (swe)/PSEN1dE9 transgenic mouse model. Urine samples were collected from the control and transgenic groups. The proteins were analyzed using liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) identification. Differential proteins were analyzed by DAVID and IPA.

Materials and methods

Experimental animals

APP (swe)/PSEN1dE9 transgenic mice and wild-type mice were used in this study. All animal protocols

governing the experiments in this study were approved by the Institute of Basic Medical Sciences

Animal Ethics Committee, Peking Union Medical College (Approved ID: ACUC-A02-2014-008). All

animals were maintained with a standard laboratory diet at a controlled indoor temperature (22 ± 1°C)

and humidity (65 ~ 70%) and with a 12-h light-dark cycle. The study was performed according to the

guidelines developed by the Institutional Animal Care and Use Committee of Peking Union Medical

College. All efforts were made to minimize suffering.

Urinary protein extraction

Urine samples from 4-month-, 6-month- and 8-month-old mice in transgenic and wide-type control

groups were collected in metabolic cages. During urine collection, all rats were given free access to

water without food to avoid contamination.

Urine was centrifuged at 2,000×g for 15 min at 4°C. After the cell debris had been removed, the

supernatant was centrifuged at 12,000×g for 15 min at 4°C. Three volumes of acetone were added after

the pellets had been removed, and precipitation was allowed to occur at 4°C. After the supernatant was

removed, the pellets were resuspended in lysis buffer (8 M urea, 2 M thiourea, 25 mM dithiothreitol

(DTT) and 50 mM Tris). Protein concentrations were measured using the Bradford method.

In-gel protein digestion

Proteins were digested with trypsin (Trypsin Gold, Mass Spec Grade, Promega, Fitchburg, Wisconsin,

USA) using in-gel protein digestion[21]. Protein samples were separated by SDS-PAGE. The gel was

rinsed three times, 5 minutes each, in pure water and stained for 30 minutes in Coomassie brilliant blue bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

at room temperature, after which the staining solution was discarded. The gel was destained for 1 hour

in pure water at room temperature. When destaining was complete, the solution was discarded. A clean

razor blade was used to cut the protein bands from the gel. The gel slices were then placed into a

0.5-ml microcentrifuge tube that had been prewashed twice with 50% acetonitrile (ACN)/0.1%

trifluoroacetic acid (TFA). The gel slices were destained twice with 0.2 ml of 100 mM NH4HCO3/50%

ACN for 45 minutes each at 37°C to remove the Coomassie brilliant blue stain. The gel slices were

dehydrated for 5 minutes at room temperature in 100 µl of 100% ACN. At this point, the gel slices

were much smaller than their original size and whitish or opaque in appearance. The gel slices were

dried in a Speed Vac® for 10 minutes at room temperature to remove the ACN. Trypsin Gold was

resuspended at 1 µg/µl in 50 mM acetic acid and then diluted in 40 mM NH4HCO3/10% ACN to 20

µg/ml. The gel slices were preincubated in a minimal volume (10–20 µl) of the trypsin solution at room

temperature (do not exceed 30°C) for 1 hour. The slices rehydrated during this time. If the gel slices

appeared white or opaque after one hour, an additional 10–20 µl of trypsin was added, and they were

incubated for another hour at room temperature. Sufficient digestion buffer (40 mM NH4HCO3/10%

ACN) to completely cover the gel slices was added, and the tubes were capped tightly to avoid

evaporation. The samples were then incubated overnight at 37°C. Next, the gel slice digests were

incubated with 150 µl of pure water for 10 minutes with frequent vortex mixing. The liquid was

removed and saved in a new microcentrifuge tube. The gel slice digests were extracted twice with 50

µl of 50% ACN/5% TFA (with mixing) for 60 minutes each time at room temperature. All extracts

were pooled and dried in a Speed Vac® at room temperature for 2–4 hours. The extracted peptides were

purified and concentrated using ZipTip® pipette tips (Millipore Corporation, Cat# ZTC18S096,

Darmstadt, Germany). The peptides eluted from the ZipTip® tips were then ready for mass

spectrometric analysis.

Filter-aided protein digestion

Proteins were digested with trypsin (Trypsin Gold, Mass Spec Grade, Promega, Fitchburg, Wisconsin,

USA) using filter-aided sample preparation methods [22]. After proteins were loaded into a 10-kDa

filter unit (Pall, Port Washington, New York, USA), UA buffer (8 M urea in 0.1 M Tris-HCl, pH 8.5)

and NH4HCO3 (25 mM) were added successively, and the tube was centrifuged at 12,000×g for 30 min

at 18°C. Proteins were denatured by incubation with 20 mM DTT at 37°C for 1 h and then alkylated bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

with 50 mM iodoacetamide (IAA) for 45 min in the dark. After the samples had been centrifuged with

UA twice and NH4HCO3 four times, the proteins were redissolved in an NH4HCO3 solution and

digested with trypsin (1:50) at 37°C overnight. The tryptic peptides were desalted using Oasis HLB

cartridges (Waters, Milford, Massachusetts, USA), and the desalted peptides were dried by vacuum

evaporation (Thermo Fisher Scientific, Bremen, Germany).

Peptide tandem mass tag (TMT) labeling

The filter-aided digested peptides were solubilized in 100 mM tetraethylammonium bromide (TEAB)

and labeled with 6-plex Tandem Mass Tag Label Reagents provided by Thermo Fisher Scientific

(Pierce, Rockford, IL, USA), which had been equilibrated to room temperature immediately before use.

Then, 41 µl of anhydrous acetonitrile was added to each tube, and the reagent was allowed to dissolve

for 5 min with occasional vortexing. The samples were briefly centrifuged to gather the solution, and

20 µl of the TMT Label Reagent was added. The reaction was then incubated for 2 h at room

temperature. After the peptides were labeled with isobaric tags, they were mixed at a 1:1:1:1:1:1 ratio

based on the amount of total peptide, which was determined by running equal volumes of labeled

samples through liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) and

comparing the total signal intensities of all peptides. Finally, samples from the wild-type group (n=3)

and AD group (n=3) were analyzed by two-dimensional LC-MS/MS (2DLC-MS/MS).

HPLC separation

The TMT-labeled samples were fractionated using a high-pH reversed-phase liquid chromatography

(RPLC) column from Waters (4.6 mm × 250 mm, Xbridge C18, 3 μm) and loaded onto the column in

buffer A1 (H2O, pH=10). The elution gradient progressed from 5–25% buffer B1 (90% ACN, pH=10;

flow rate=1 ml/min) over 60 min. The eluted peptides were collected at one fraction per minute. The 60

fractions were dried, resuspended in 0.1% formic acid and pooled into 30 samples by combining

fractions 1 and 31, 2 and 32, and so on. The odd-numbered fractions were chosen for further analysis.

A total of 15 fractions from urinary peptide mixtures were analyzed by LC-MS/MS.

LC-MS/MS analysis bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

The in-gel digested peptides and TMT-labeled peptides were analyzed with a reverse-phase C18

self-packed capillary LC column (75 μm × 100 mm). The elution gradient progressed from 5–30%

buffer B2 (0.1% formic acid, 99.9% ACN; flow rate=0.3 μL/min) over 40 min. A Triple TOF 5600

mass spectrometer was used to analyze the peptides eluted from LC, and each fraction was run twice.

MS data were acquired using the high-sensitivity mode with the following parameters: 30

data-dependent MS/MS scans per full scan, full scans acquired at a resolution of 40,000, MS/MS scans

acquired at a resolution of 20,000, rolling collision energy, charge state screening (including precursors

with a charge state of +2 to +4), dynamic exclusion (exclusion duration 15 s), an MS/MS scan range of

100-1800 m/z, and a scan time of 100 ms.

Data analysis

The MS/MS data were analyzed with Mascot software (version 2.4.1, Matrix Science, London, UK),

and proteins were identified by comparing the peptide spectra against the Swissprot_2014_07

databases (taxonomy: mus). Trypsin was selected as the digestion enzyme, up to two missed cleavage

sites were allowed, and carbamidomethylation of a cysteine was defined as a fixed modification. The

precursor ion mass tolerance and the fragment ion mass tolerance were 0.05 Da. The TMT-labeled

protein identification Mascot results were validated by using Scaffold Proteome Software (version

Scaffold_4.3.3, Proteome Software Inc., Portland, OR). Peptide identification was accepted at a false

discovery rate (FDR) of less than 1.0% at the protein level and if the sample produced at least 2 unique

peptides. The Scaffold Q+ was used for Label-Based Quantification (TMT, iTRAQ, SILAC, etc.) of

peptides and proteins. Normalized reporter ion intensities were used to calculate the relative protein

abundance. The protein ratios were then quantified by the median of the transformed reporter ion

intensity ratios [23]. Label-free quantification was performed using Progenesis LC-MS software

(version 4.1, Nonlinear, UK) as described previously [24].

Statistical analysis

All statistical analysis was performed with the Statistical Package for Social Studies software (SPSS,

version 16, IBM). Comparisons between independent groups were conducted using one-way ANOVA

followed by post hoc analysis with the least significant difference (LSD) test, and P-values less than

0.05 were considered to indicate statistically significant differences. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Results and discussion

Urinary proteome changes in APP (swe)/PSEN1dE9 transgenic mice

The urinary proteins of 4-month-old mice were digested by an in-gel method and quantified by a

label-free method. A total number of 443 protein groups were identified in the urinary proteome at an

FDR < 1%, and each generated at least 2 unique peptides. The levels of 33 urinary proteins were

significantly different between the AD and wild-type groups (Table 1); among these proteins,

cadherin-16 levels were different in both the >85 kDa and 50-85 kDa lanes, sulfhydryl oxidase 1 levels

were different in both the 50-80 kDa and 30-50 kDa lanes, and deoxyribonuclease-1 levels were

different in both the 15-30 kDa and <15 kDa lanes. The fold change of the significantly changed

protein levels was more than 1.5. In these 33 differential proteins, 17 had been reported to be

associated with AD. Specifically, 15 had been reported to be associated with the pathology

mechanisms of AD, while 9 were identified as direct AD biomarkers (Table 1).

Seven proteins relate to the development of AD as well as biomarkers of AD. (1) Kallikrein-1 (KLK1)

levels were different in 4-, 6- and 8-month-old groups than the wild-type group, and the

kallikrein-kinin system mediates inflammation in AD in vivo[25]. Kallikrein-6, one member of the

kallikrein family, was reported as a biomarker of AD and to have increased levels in CSF and serum

and decreased levels in tissue [26]. (2) Galactocerebrosidase (GALC) levels were different in 4- and

8-month-old groups than the wild-type group, and a deficiency in galactosylceramidase potentially

contributed to neurodegenerative disease [27]. Ceramide has been suggested to participate in the

neuronal cell death that leads to AD. GALC, as a gene connected to ceramide metabolism, was

upregulated in the brain tissue of AD patients, making it an attractive candidate for diagnostic purposes

and for intervening in neurodegenerative processes [28]. (3) Ceruloplasmin (CERU) levels were higher

in both 4- and 8-month-old groups than in the control group, with the same trend of increased levels

found in the serum of AD patients[29]. The levels of CSF diagnostic markers, such as Aβ42, tau and

phospho-tau, were correlated with lower plasma copper and CERU levels in patients with

Alzheimer’s disease[30]. The ratio of CERU concentrations measured by enzymatic methods (eCP) to

those measured by immunological methods (iCP), eCP/iCP, reflects the high specificity of AD patients

as well as a decreased risk of having AD[31]. CERU had less ferroxidase activity in AD patients than

wild-type patients, which contributed to the development of AD[32]. (4) Fibronectin (FINC) levels bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

were different in 4- and 8-month-old groups than in the wild-type group and were reported to be a

novel biomarker for AD from blood[33]. In addition, FINC levels were significantly lower in plasma in

mild cognitive impairment (MCI) patients than healthy patients, which provides further insight into

the biological pathways and processes that underpin the pathophysiology and progression of MCI and

AD[34]. (5) Cathepsin B (CATB) levels were different in 4- and 8-month-old groups than the

wild-type group. CATB has been previously reported to be upregulated in brain tissues from APP/PS1

transgenic mice, and its levels changed in the same direction relative to healthy mice as they changed

in the serum of AD patients relative to control patients, making CATB a potential biomarker of AD[13].

In addition, CATB produces brain pyroglutamate amyloid beta, which represents a potential AD

therapeutic[35]. (6) Angiotensinogen (ANGT) was upregulated in 4-month-old groups and has

exhibited increased levels in the CSF of AD patients[36]. As a component of the renin-ANGT system

(RAS), ANGT is helpful for AD processes[37]. (7) Annexin A11 (ANX11) was upregulated in

4-month-old groups. Annexin A1 was reported to be expressed strongly in the microglia of AD patients,

which aids the surveillance of microglia and the maintenance of brain homeostasis by using Annexin

A1-dependent mechanisms[38], and Annexin A5 was reported to be a biomarker of AD and present at

increased levels in the plasma of AD patients[39]. The protein-normalized abundances and spectral

counts of all seven differential proteins were greater in every mouse in the high-abundance group than

those in the low-abundance group.

Two other differential proteins have been mentioned as potential AD biomarkers. (1) Ectonucleotide

pyrophosphatase/phosphodiesterase family member 2 (ENPP2) levels were higher in the CSF of AD

patients than the CSF of healthy controls. ENPP2 can be used to specifically discriminate AD from

Lewy body dementia, making it a candidate AD biomarker[40]. (2) The Ig kappa chain C (IGKC)

region was reported to be a potential serum biomarker of AD at the early stage[41].

Eight differential proteins are associated with the mechanism of AD pathology. (1) CDH1, the gene of

cadherin-1, is an AD risk gene that has significant association in Caribbean Hispanic individuals[42].

(2) As glycosylated transmembrane proteins, cadherins such as cadherin-1 and cadherin-16 also

directly interact with the presenilin-1/g-secretase complex, which contributes to the modulation of Aβ

peptide formation and thus plays a critical role in AD etiology[43]. (3) Alpha-amylase 1 (AMY1)

reflects M3 muscarinic agonist activity in AD[44]. (4) ASMase, the gene of sphingomyelin bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

phosphodiesterase, which is connected to ceramide metabolism, was upregulated in the brain tissue of

AD patients, making it an attractive candidate for both diagnostic purposes and intervening in

neurodegenerative processes[28]. Significant correlations were observed between brain sphingomyelin

phosphodiesterase levels and the levels of amyloid beta peptide (Aβ) and phosphorylated tau protein,

which contribute to AD disease pathogenesis[45]. (5) Platelet-activating factor acetylhydrolase (PAFA)

was present in higher levels in the plasma of AD patients than that of healthy controls, and these higher

levels were associated with oxidative damage of low-density lipoproteins (ox-LDL), which contributes

to the inflammation and oxidative stress of plasma lipoproteins and is strongly associated with

Alzheimer’s disease[46]. (6) Plasma carbonic anhydrase 2 (CAH2) was elevated in AD patients and

may thus play a role in the pathogenesis of AD[47]. (7) Acid ceramidase (ASAH1) levels were

elevated in AD brain, suggesting that acid ceramidase might play a role in controlling neuronal

apoptosis and that acid ceramidase-mediated signaling pathways might be involved in the molecular

mechanism of AD[48]. (8) Lipoprotein lipase (LIPL) is associated with neurite pathology, and its levels

were reduced in the dentate gyrus of AD brains[49]. Some differential proteins have not been reported

in research papers and may provide clues to candidates involved in the AD pathology mechanism.

Figure 3. Urinary proteins with significantly different levels in the 4-, 6- and 8-month-old mice and

wild-type mice

The urinary proteins of 6-month-old mice were digested by a filter-aided method and quantified by a

TMT-labeling method. At the protein level, 310 protein groups were identified in the urinary proteome

at an FDR < 1%, and each generated at least 2 unique peptides. The levels of 33 urinary proteins were

significantly different between the AD and wild-type groups, and their fold changes were greater than bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1.5. Eight of these 33 differential proteins had been reported to be associated with AD. To be specific,

6 were associated with the pathology mechanisms of AD, while 3 had been directly referred to be AD

biomarkers (Table 2).

Three differential proteins (KLK1, IGKC, and RAB10) have been suggested as AD biomarkers. The

target gene RAB10, which is regulated by miRNA-369-3p, miRNA-30e-5p, miRNA-30e-3p, and

miRNA-655, may play key roles in the progression and development of AD and may be a biomarker of

AD[50].

Six differential proteins are associated with the mechanism of AD pathology. (1) Neogenin (NEO1)

expresses weak NEO1 immunoreactivity in a small subset of amyloid plaques of AD brains[51]. (2)

Charged multivesicular body protein 2a (CHM2A) levels were elevated in the 6-month-old group.

Mutations in charged multivesicular body protein 2B (CHMP2B) can give rise to FTLD-UPS[52]. (3)

Peroxisomal carnitine (OCTC) levels were higher in the 6-month-old group than the wild-type group,

and new therapies acting on peroxisomal metabolism might be developed to prevent cognitive decline

and other age-related neurological disorders[53]. (4) Heat shock cognate 71 kDa protein (HSP7C) was

upregulated in 6-month-old groups, while its gene level was significantly lower across the three brain

regions in AD patients than in controls patients, suggesting their participation in AD

pathogenesis[54]. (5) Cooperative expression of leukemia inhibitory factor (LIF) and its receptor

(LIFR) in AD hippocampus may indicate a role for LIF in neuronal damage or repair at these sites[55].

The urinary proteins of 8-month-old mice were digested by an in-gel method and quantified by a

label-free method. At the protein level, 273 proteins were identified in the urinary proteome at an FDR

< 1%, and each generated at least 2 unique peptides. The levels of 89 urinary proteins were

significantly different between the AD and wild-type groups. Thirty-three of these 89 differential

proteins have been associated with AD. Specifically, 29 are associated with the mechanism of AD

pathology, while 17 have been suggested as candidate AD biomarkers. (Table 3)

Thirteen proteins have been previously reported to be related to the development of AD as well as

biomarkers of AD, while 5 of them (KLK1, GALC, CERU, FINC, and CATB) had been already noted

in 4-month-old groups. For example, (1) Serum albumin (ALBU) was differentially expressed between

brains of AD mouse models and those of healthy mice[56], and serum levels of albumin-amyloid beta bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

complexes were lower in AD patients than healthy controls, which is very useful to apply in

monitoring the progression of AD[57]. (2) Clusterin (CLUS) levels were higher in AD patients than

controls, especially in regions with most abundant in Aβ, facilitating the development of AD[58]. As a

heterodimeric glycoprotein, clusterin was more abundant in the CSF of some neurodegenerative

disease patients than that of healthy controls and could thus serve as a potential biomarker to

differentiate Parkinson’s disease (PD) from dementia with Lewy bodies (DLB)[59]. (3) CSF

complement C3 (CO3) is a staging biomarker in AD[60]. In addition, the dysregulation of neuron-glia

interactions through NFB/ C3/C3aR signaling may contribute to synaptic dysfunction in AD[61].

Seventeen differential proteins have been mentioned as potential AD biomarkers. For example, (1)

sulfated glycoprotein 1 (SAP) is a novel CSF biomarker for neurodegeneration and has been validated

by a high-throughput multiplexed targeted proteomic assay[40]. (2) Cystatin-C (CYTC) is a novel CSF

biomarker for staging early AD[62]. Twenty-nine differential proteins are associated with the

mechanism of AD pathology. For example, (1) CD44 antigen (CD44) gene expression was greater in

lymphocytes derived from AD patients than in those from healthy controls, suggesting that it plays

roles in the peripheral immune response during the development of AD[63]. (2) The rs2227564

polymorphism of the urokinase-type plasminogen activator (UROK) gene increased the risk of AD[64].

Other differential proteins are annotated in Table 3.

Functional analysis of differential urine proteins in APP (swe)/PSEN1dE9 transgenic mouse

4 month

lipid metabolic process

lipid catabolic process

metabolic process

calcium-dependent cell-cell adhesion via plasma membrane cell adhesion molecules

decidualization

carbohydrate metabolic process

0 1 2 3 4 -log(P-value) bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

6 month

microvillus assembly

regulation of cell shape

establishment of protein localization to plasma membrane

terminal web assembly

intestinal D-glucose absorption

regulation of microvillus length

transport

positive regulation of establishment of protein localization to plasma membrane

0 1 2 3 4 5 -log (P-value)

8 month

proteolysis metabolic process carbohydrate metabolic process proteolysis involved in cellular protein catabolic process fibrinolysis peptide catabolic process blood coagulation acute-phase response cell adhesion lysosome organization biotin metabolic process chaperone-mediated autophagy angiogenesis cellular iron ion homeostasis positive regulation of cell proliferation positive regulation of apoptotic cell clearance organ regeneration regulation of smooth muscle cell migration glycosaminoglycan metabolic process negative regulation of fibrinolysis cell-matrix adhesion complement activation, alternative pathway glycosphingolipid metabolic process central nervous system development

0 2 4 6 8 0 1 -log(P-value)

Figure 4. Biological processes of differential proteins in different stages of AD development. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Functional annotation of differential urinary proteins was performed using DAVID[65]. The

differential proteins at three time points were classified to be involved with certain biological processes

(Figure 4), molecular components, and molecular functions (Supplemental Figures 1 and 2).

In 4-month-old mice, differential urine proteins involved in lipid metabolism and lipid catabolism were

enriched. As lipid metabolism is a fundamental process for brain development and function, aberrant

lipid metabolism was not surprisingly common in an animal model of AD that relates to AD

pathogenesis[66-68]. For example, AD-induced perturbation of niche fatty acid metabolism can

suppress the homeostatic and regenerative functions of neural stem cells, supporting the mechanism of

AD pathogenesis[66]. Abnormal lipid metabolism influences Aβ metabolism and deposition in both

brain parenchyma and vasculature, as well as tau hyperphosphorylation and aggregation, which then

likely triggers a series of downstream catalytic events that eventually affect the progression of AD

pathogenesis[67]. More importantly, lipid rafts from human cerebral cortex are associated with the

pathogenesis of early AD, as -secretase/APP (amyloid-protein precursor) interactions and lipid raft

microviscosity are strongly and positively correlated in AD frontal and entorhinal cortices, indicating

that the aberrant lipid metabolism had already occurred in the early stage of AD[69]. Decidualization

and calcium-dependent cell-cell adhesion via plasma membrane cell adhesion molecules, which are not

reported in AD studies, may play roles in the mechanism of AD pathology. Recent studies

demonstrated that using a triple receptor agonist (TA), which activates GIP-1, GIP and glucagon

receptors at the same time, reduces the total amount of beta amyloid in an APP/PS1 transgenic mouse

model through a 2-month TA treatment [70]. However, this research used 6-month-old transgenic mice

for a 2-month treatment, and beta amyloid had already appeared. We thus suppose that a more effective

treatment can be achieved when the 4-month-old transgenic mice are used for therapy.

The differential proteins in 6-month-old mice were involved in transport processes, while axonal

transport (AT) defects play an important role in the pathogenesis of AD[71]. Other biological processes

such as microvillus assembly, the regulation of cell shape and protein localization to the plasma

membrane may contribute to the formation of beta amyloid plaque deposits in the brain of AD patients,

which provides potential clues for the studying the mechanism of AD.

A larger number of differential urine proteins were identified in 8-month-old mice and indicated other

AD-related biological processes. Proteolysis processes were overrepresented and reported to be bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

associated with APP proteolysis, which is consistent with the pathology of AD[72]. In addition, the

urinary proteins that were related to fibrinolysis may be factors contributing to AD[73]. What’s more,

lysosome organization accumulated at amyloid plaques in mouse models of AD, which gave rise to

potential therapies of AD patients[74]. Lastly, angiogenesis processes were reported to contribute to

the pathogenesis of AD, as they are a common feature of amyloid beta (Aβ) plaques[75, 76], and the

levels of some angiogenesis factors were elevated in the plasma of AD patients[77].

In summary, in the biological process category, metabolic and carbohydrate metabolic processes were

overrepresented in 4-and 8-month-old AD mice. The differential urine proteins in 4- and 8-month-old

mice indicated an enrichment of metabolic process, while the three subtypes of AD can be

distinguished by metabolic profiling[78]. Considering that the biological processes whose activities

changed in this experiment did not extensively overlap at these three time points, this result indicates

that only three different stages might occur, implying that different therapeutic strategies may be

appropriate for different stages. If biomarkers can be found in human as early as was found for the

4-month-old in this model, we may have an early intervention window in which disease development

can be stopped.

In the cellular component category (Supplemental Figure 1), most differential proteins came from the

extracellular exosome, the extracellular space, the anchored component of membranes, and blood

microparticles, whereas a small number of differential proteins were derived from organelles such as

the lysosome and the Golgi apparatus. In the molecular function category (Supplemental Figure 2),

hydrolase activity was overrepresented in 4- and 8-month-old mice; calcium ion binding was

overrepresented in 4-month-old mice; protein complex binding, sodium ion transmembrane transporter

activity and protein binding were overrepresented in 6-month-old mice; and peptidase, endopeptidase,

serine-type peptidase activities were overrepresented in 8-month-old mice.

IPA of differential proteins in AD development

To identify the differential proteins with mostly affected cellular death and survival, ingenuity pathway

analysis (IPA) was applied to cell death and survival networks (Figure 5). Cadherin-1, which was

reported to be associated with AD[42], is a core component of these networks in 4-month-old mice.

CDH16, CP, ENPP2, AGT, SMPD1, PLA2G7 were upregulated and play roles in AD[28, 29, 36, 40, bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

43, 46], whereas sialate O-acetylesterase was downregulated in 4-month-old mice. In 6-month-old

mice, NEO1, CROT, GP2, SEMA4A and LIFR were present in higher levels in transgenic mice than

the control group, and NEO1 as well as CROT were associated with the pathology and mechanism of

AD[51, 79]. In contrast, MEP1B, CLIC4, and PDZK1 levels were lower in transgenic mice than the

control group. FINC was the core component of the network in 8-month-old mice and plays roles in

AD[80]. In addition, AFM, MFGE8, C3, CD44 and PLAU were upregulated in transgenic mice and are

potential biomarkers or play roles in the development of AD[61, 63, 64, 81, 82], while VNN1, HEXB

and CTSZ were downregulated in transgenic mice when compared with control groups.

4-month-old 6-month-old 8-month-old

Figure 5. IPA of differential proteins in AD development

Conclusion

Our results indicated that urine proteins enable AD early detection before beta amyloid plaque

deposition, which may provide an opportunity for intervention.

Acknowledgements

This research was supported by the National Key Research and Development Program of China (2016

YFC 1306300), Key Basic Research Program of the Ministry of Science and Technology of China

(2013FY114100), Beijing Natural Science Foundation (7173264, 7172076), Beijing cooperative

construction project (110651103）Beijing Normal University (11100704) Peking Union Medical

College Hospital (2016-2.27)，the National Natural Science Foundation of China (NSFC #81271239,

#81771205, #91632113), the Natural Science Foundation and Major Basic Research Program of

Shanghai (16JC1420500, 16JC1420502), and the CAMS Innovation Fund for Medical Sciences bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

(CIFMS #2017-I2M-3-008). The funders had no role in study design, data collection and analysis,

decision to publish, or preparation of the manuscript. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Table 1. Details of differential urinary proteins of 4-month-old mice Normalized abundance pathology biom Human P Fold Confidence MW Protein name UniProt and arker UniProt value change score WT-1 WT-2 WT-3 AD-1 AD-2 AD-3 mechanism s Cadherin-1 P09803 P12830 0.022 1.87 218.7 2619939 3821894 4545058 7387696 5702665 7448193 [42, 43] - >85 Cadherin-16 O88338 O75309 0.039 1.33 475.3 7842297 10683716 8044870 12305698 10097708 12879662 [43] - kDa Keratin, type II Q3UV1 Q01546 0.041 0.56 181.9 95619 152124 160984 52841 108167 67233 - - cytoskeletal 2 oral 7 Ceruloplasmin Q61147 P00450 0.018 3.19 142.8 1405128 700799 1034296 4233275 2979048 2817111 [31, 32] [30] Alpha-amylase1 P00687 P04745 0.037 2.98 1271.6 16862562 16394995 16747254 27733747 36333676 85064392 [44] - Ectonucleotide Q9R1E6 Q13822 0.038 2.89 438.4 6410477 4227746 3261439 14447745 15522911 10137661 - [40] pyrophosphatase AnnexinA11 P97384 P50995 0.035 2.54 124.5 112847 385493 224516 851694 561465 420834 [38] [39] Angiotensinogen P11859 P01019 0.036 2.53 340.3 1832470 6008159 4047815 8130284 9777242 12155633 [37] [36] Kallikrein-1 P15947 P06870 0.028 2.55 161.4 13993666 15481983 8200572 32915114 24952991 38216076 [25] [26] Q9Z0K Pantetheinase O95497 0.010 2.25 329.7 7264591 9291049 8941437 18429885 12901960 26056688 - - 8 50-85 Q8BND Sulfhydryloxidase1 O00391 0.029 1.93 444.1 4216654 7066885 5327850 12135722 10500098 9505950 - - kDa 5 Sphingomyelin Q04519 P17405 0.004 1.71 138.9 789408 934631 809095 1433994 1281043 1625436 [28, 45, 83] - phosphodiesterase N-Acetylmuramoyl-L- Q8VCS Q96PD5 0.001 1.63 237.3 4431595 4089554 4317879 6991718 6218791 7761766 - - alanine amidase 0 Galactocerebrosidase P54818 P54803 0.002 1.51 296.1 23253378 21116228 25380610 39727782 31492109 33948684 [27] [28] Cadherin-16 O88338 O75309 0.005 1.37 165.0 8088405 10041828 9001115 12089782 12239425 12726177 [43] - Trehalase Q9JLT2 O43280 0.036 1.38 176.5 1708628 1343343 1240247 2367601 1874206 1677882 - - Platelet-activating Q60963 Q13093 0.024 1.49 103.9 790161 1089782 1129603 1525671 1594208 1369456 [46] - factor acetylhydrolase Carbonic anhydrase 2 P00920 P00918 0.050 7.64 139.9 1026487 137540 1196867 2483019 8183767 7363671 [47] - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

ATP-binding cassette Q86UQ sub-family A member Q5SSE9 0.003 3.97 186.0 162885 413064 239944 1326933 1045477 870531 - - 4 13 Protein LEG1 Q8C6C Q6P5S2 0.046 3.17 331.5 617434 5666756 2910064 10253781 9689764 9179997 - - homolog 9 Cathepsin B P10605 P07858 0.029 3.13 338.5 7139961 1114522 1317986 9277618 8644600 12074430 [35] [13] Q9WV5 Acid ceramidase Q13510 0.031 0.42 264.6 39183681 22168743 30757070 11403664 10956829 16671159 [48] - 30-50 4 kDa Q8BND Sulfhydryl oxidase 1 O00391 0.005 0.40 271.4 6515045 6236645 3506111 2174726 1897719 2494845 - - 5 Eosinophil cationic 25123514 10362902 27345839 11217001 P97426 no 0.031 0.39 167.5 66736059 65435575 - - protein 1 5 6 6 0 Fibronectin P11276 P02751 0.017 0.36 191.7 6815535 4157474 3089723 1594554 2037884 1485064 [34] [33] Prostaglandin-H2 11633843 11891642 O09114 P41222 0.018 0.27 196.1 35968640 25071521 28557837 19099588 - - D-isomerase 5 2 Q9UHL Dipeptidyl peptidase 2 Q9ET22 0.017 0.23 329.3 36130176 10831853 12769425 4129725 4626310 4946023 - - 4 Lipoprotein lipase P11152 P06858 0.025 0.18 157.9 12700373 4762033 4358911 1006139 706457 2137022 [49] - Sialate Q9HAT P70665 0.041 0.17 150.5 5186756 1470459 629510 242721 323999 669184 - - O-acetylesterase 2 Alpha-1-antitrypsin Q00897 no 0.029 0.09 386.0 47453084 7866154 10014171 1352876 1413734 3276721 - - 1-4 Deoxyribonuclease-1 P49183 P24855 0.002 2.46 177.0 6661598 5715618 7105781 11646345 22351861 13973828 - - 15-30 Ig kappa chain C kDa P01837 P01834 0.014 1.80 103.9 8822671 11713287 12329617 21590819 21713190 15916672 - [41] region Deoxyribonuclease-1 P49183 P24855 0.026 2.47 146.4 1360757 728821 1080584 1869443 3272391 2687073 - - <15 Secretoglobin family kDa Q9JI02 no 0.026 0.56 323.6 15981747 29749962 27032579 15560357 9828085 15548114 - - 2B member 20 bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Ig kappa chain V-V P01642 no 0.021 0.20 130.7 893023 1868961 641920 208754 175548 309125 - - region L7 (Fragment)

Table 2. Details of differential urinary proteins of 6-month-old mice

Normalized ratio Human P Fold Spectral pathology and Protein name UniProt MW biomarker UniProt value change count mechanism WT-1 WT-2 WT-3 AD-1 AD-2 AD-3

Major urinary protein 1 P11588 NO 0.004 10.92 1657 21 kDa 0.45 0.95 1.6 8.45 13.9 10.4 - - Neogenin P97798 Q92859 0.005 7.22 3 38 kDa 1 1.45 0.1 4.8 7.75 5.85 [51] - CMRF35-like molecule Q8K249 Q496F6 0.004 5.31 3 22 kDa 0.5 3 3.4 12.45 9.7 14.5 - - 2 Major urinary protein 5 P11591 no 0.011 3.79 326 21 kDa 1.05 4 3 12.45 7.55 10.5 - - WAP four-disulfide Q9JHY Q8WWY7 0.029 2.94 5 10 kDa 0.9 0.95 0.65 3.25 2.5 1.6 - - core domain protein 12 3 Keratin, type I P19001 P08727 0.014 2.41 9 45 kDa 1 1.2 1.25 3.35 2.05 2.9 - - cytoskeletal 19 Cadherin-15 P33146 P55291 0.015 2.06 3 86 kDa 1 0.6 0.75 1.95 1.4 1.5 - - Charged multivesicular Q9DB3 O43633 0.027 1.88 4 25 kDa 1 0.8 0.65 1.35 1.35 1.9 [52] - body protein 2a 4 Cadherin-11 P55288 P55287 0.024 1.87 2 88 kDa 1 0.7 0.55 1.65 1.25 1.3 - - Q9DC5 Peroxisomal carnitine Q9UKG9 0.030 1.86 2 70 kDa 1 1.05 0.85 2.25 1.75 1.4 [53] - 0 Desmoplakin E9Q557 P15924 0.041 1.80 27 333 kDa 0.8 0.85 0.8 1.3 1.2 1.9 - - Pancreatic secretory granule membrane Q9D733 P55259 0.048 1.79 3 59 kDa 1 0.8 0.55 1.55 1.05 1.6 - - major glycoprotein GP2 bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Alpha-1-antitrypsin 1-4 Q00897 no 0.022 1.75 6 46 kDa 0.9 0.8 0.9 1.45 1.85 1.25 - - Heat shock cognate 71 P63017 P11142 0.020 1.67 7 71 kDa 0.9 0.9 0.95 1.45 1.3 1.85 [54] - kDa protein Kynurenine/alpha-amin oadipate Q9WV Q8N5Z0 0.002 1.63 2 48 kDa 1 1.15 0.85 1.65 1.6 1.65 - - aminotransferase, M8 mitochondrial Semaphorin-4A Q62178 Q9H3S1 0.002 1.62 5 83 kDa 0.85 0.7 0.7 1.2 1.15 1.3 - - Serine protease Q03734 no 0.006 1.53 3 47 kDa 0.9 0.95 0.8 1.2 1.45 1.4 - - inhibitor A3M Leukemia inhibitory P42703 P42702 0.044 1.51 6 122 kDa 0.9 0.9 1.05 1.45 1.15 1.7 [55] - factor receptor Kallikrein-1 P15947 P06870 0.028 0.61 79 29 kDa 1 1.5 1.35 0.8 0.75 0.8 [25] [26] Prominin-1 O54990 O43490 0.027 0.60 2 97 kDa 1 0.75 0.75 0.55 0.4 0.55 - - Glycerophosphodiester phosphodiesterase Q99LY Q7L5L3 0.039 0.59 7 38 kDa 0.9 1.35 1.4 0.65 0.7 0.8 - - domain-containing 2 protein 3 Ig kappa chain C region P01837 P01834 0.003 0.57 25 12 kDa 1 0.85 0.8 0.5 0.5 0.5 - [41] Solute carrier family 12 P59158 P55017 0.030 0.55 3 111 kDa 1 0.9 1.3 0.5 0.75 0.5 - - member 3 Meprin A subunit beta Q61847 Q16820 0.029 0.54 6 79 kDa 1.1 1.05 0.9 0.3 0.7 0.65 - - Solute carrier family 23 Q9Z2J0 Q9UHI7 0.014 0.54 9 66 kDa 1 0.7 0.8 0.4 0.45 0.5 - - member 1 Sushi Q9DBX domain-containing no 0.001 0.54 3 90 kDa 0.9 1 0.9 0.5 0.45 0.55 - - 3 protein 2 Villin-1 Q62468 P09327 0.011 0.52 8 93 kDa 1 0.7 0.8 0.4 0.45 0.45 - - Ezrin P26040 P15311 0.027 0.51 11 69 kDa 0.9 0.55 0.7 0.35 0.4 0.35 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Chloride intracellular Q9QYB Q9Y696 0.008 0.50 3 29 kDa 1 0.7 0.9 0.45 0.45 0.4 - - channel protein 4 1 Na(+)/H(+) exchange regulatory cofactor Q9JIL4 Q5T2W1 0.013 0.48 4 56 kDa 1 0.8 0.7 0.3 0.45 0.45 - - NHE-RF3 Ras-related protein P61027 P61026 0.041 0.43 2 23 kDa 1 1.05 1.65 0.4 0.75 0.45 - [50] Rab-10 Na(+)/H(+) exchange regulatory cofactor P70441 O14745 0.011 0.42 13 22 kDa 1 0.65 0.75 0.3 0.35 0.35 - - NHE-RF1 Ig kappa chain V-V P01642 no 0.005 0.38 10 13 kDa 1 0.8 0.7 0.35 0.3 0.3 - - region L7 (Fragment)

Table 3. Details of differential urinary proteins of 8-month-old mice Spectral count pathology biomarke p and r MW range Protein name UniProt Human UniProt MW W WT WT AD AD AD value mechanis T1 2 3 1 2 3 m Angiotensin-converting enzyme P09470 P12821 0.0092 151 kDa 27 28 27 20 13 19 [84] - Aminopeptidase N P97449 P15144 0.025 110 kDa 6 4 8 15 10 14 [85] - >100 kDa Uromodulin Q91X17 P07911 0.027 71 kDa 95 98 165 186 208 227 - - Lysosome-associated membrane glycoprotein 2 P17047 P13473 0.031 46 kDa 3 2 2 1 0 1 [86] [87] Podocalyxin Q9R0M4 O00592 0.042 53 kDa 0 0 0 2 1 4 - - Beta-glucuronidase P12265 P08236 0.0001 74 kDa 5 4 6 0 0 0 [88] - 0.0002 - - 65-100 Carboxylesterase 1C P23953 no 61 kDa 17 18 17 12 12 11 9 kDa Biotinidase Q8CIF4 P43251 0.0034 58 kDa 2 2 3 0 0 0 - - Major urinary protein 20 Q5FW60 no 0.009 21 kDa 7 7 6 4 5 5 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Galactocerebrosidase P54818 P54803 0.011 77 kDa 1 2 2 0 0 0 [27] [28] Pancreatic alpha-amylase P00688 no 0.016 57 kDa 0 0 0 6 4 10 - - Uromodulin Q91X17 P07911 0.016 71 kDa 39 33 43 75 81 115 - - Complement factor D P03953 P00746 0.016 28 kDa 1 0 2 8 4 5 - - Tyrosine-protein kinase receptor UF Q00993 P30530 0.018 98 kDa 0 0 1 2 2 2 - - 21 [57] [56] Serum albumin P07724 P02768 0.02 69 kDa 187 174 153 137 127 5 Isoform 2 of V-type proton ATPase catalytic subunit - [56] P50516 P38606 0.02 56 kDa 1 1 2 0 0 0 A Pantetheinase Q9Z0K8 O95497 0.021 57 kDa 2 2 2 2 1 1 - - Ceruloplasmin Q61147 P00450 0.036 121 kDa 1 1 0 2 2 2 [31, 32] [30] Alpha-N-acetylglucosaminidase O88325 P54802 0.041 ? 8 10 12 2 1 7 - - Meprin A subunit alpha P28825 Q16819 0.042 84 kDa 33 35 31 40 43 36 [89] - Hemopexin Q91X72 P02790 0.042 51 kDa 18 14 13 21 24 19 [90] [91] Ig mu chain C region P01872 no 0.049 50 kDa 0 0 0 3 2 7 - - Glycosylation-dependent cell adhesion molecule 1 Q02596 no 0.0002 16 kDa 0 1 0 8 10 10 - - 0.0004 - [81] Afamin O89020 P43652 69 kDa 1 1 0 6 7 6 9 Protein Fcgbp E9Q0B5 no 0.0005 275 kDa 2 4 2 10 12 10 - - 0.0007 - - Alpha-2-macroglobulin Q61838 no 166 kDa 11 8 9 0 0 0 6 0.0009 - - 50-65 kDa N-acetylgalactosamine-6-sulfatase Q571E4 P34059 58 kDa 11 15 11 0 0 0 6 Lysosomal Pro-X carboxypeptidase Q7TMR0 P42785 0.0013 55 kDa 8 7 5 0 0 0 - - Beta-hexosaminidase subunit beta P20060 P07686 0.0021 61 kDa 15 12 20 0 0 1 - - Pancreatic alpha-amylase P00688 no 0.0024 57 kDa 57 48 42 82 79 80 - - Aminopeptidase N P97449 P15144 0.0025 110 kDa 0 0 0 4 5 6 [85] [92] Growth arrest-specific protein 1 Q01721 P54826 0.0026 36 kDa 1 0 0 4 4 5 [93] - Desmocollin-2 P55292 Q02487 0.0027 100 kDa 0 0 0 2 3 2 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Acidic mammalian chitinase Q91XA9 Q9BZP6 0.0029 52 kDa 1 0 0 14 11 8 - - Beta-2-glycoprotein 1 Q01339 P02749 0.0038 39 kDa 1 0 0 5 4 6 - - Major urinary protein 20 Q5FW60 NO 0.0047 21 kDa 11 11 9 6 6 6 - - Serotransferrin Q921I1 P02787 0.0062 77 kDa 18 12 6 32 30 29 [94] [95] Acyloxyacyl hydrolase O35298 P28039 0.0073 65 kDa 3 2 4 0 0 0 - - Alpha-N-acetylgalactosaminidase Q9QWR8 P17050 0.0096 47 kDa 2 2 1 0 0 0 - - Lysosomal thioesterase PPT2 O35448 Q9UMR5 0.0098 34 kDa 1 2 1 0 0 0 - - Protein LEG1 homolog Q8C6C9 Q6P5S2 0.011 38 kDa 7 12 10 1 3 1 - - A0A075B5 - Protein Ighg2b no 0.011 36 kDa 0 0 0 7 3 6 - P3 Acid sphingomyelinase-like phosphodiesterase 3a P70158 Q92484 0.011 50 kDa 7 5 5 3 1 2 - - Kallikrein-1 P15947 P06870 0.012 29 kDa 2 4 5 8 7 9 [25] [26] Lactadherin P21956 Q08431 0.012 51 kDa 1 3 1 8 6 5 [82] - Alpha-1-antitrypsicn 1-5 Q00898 P01009 0.015 46 kDa 22 17 14 7 8 7 - [96] Protein 9530053A07Rik E9PVG8 Q9Y6R7 0.015 ? 26 11 24 0 1 0 - - Sushi domain-containing protein 2 Q9DBX3 Q9UGT4 0.016 91 kDa 2 4 4 0 1 1 - - Polymeric immunoglobulin receptor O70570 P01833 0.017 85 kDa 8 10 13 16 21 20 - - Protein Ighg2c F6TQW2 no 0.018 ? 0 1 2 4 5 3 - - Carboxylesterase 1C P23953 no 0.018 61 kDa 1 2 0 4 3 3 - - Tyrosine-protein kinase receptor UF Q00993 P30530 0.018 98 kDa 0 0 0 1 2 1 - - Hemopexin Q91X72 P02790 0.019 51 kDa 4 5 0 11 11 18 [90] [91] Tripeptidyl-peptidase 1 O89023 O14773 0.019 61 kDa 1 3 3 0 0 0 - - Napsin-A O09043 O96009 0.02 46 kDa 12 11 11 17 23 26 [97] - Complement C3 P01027 P01024 0.02 186 kDa 0 0 0 6 2 4 [61, 98] [60] Galactocerebrosidase P54818 P54803 0.021 77 kDa 3 6 5 1 2 2 [27] [28]- Matrix-remodeling-associated protein 8 Q9DBV4 Q9BRK3 0.023 50 kDa 0 1 0 5 3 2 - - CD44 antigen P15379 P16070 0.027 86 kDa 2 2 2 3 4 3 [63] - Major urinary protein 2 P11589 no 0.031 21 kDa 63 70 53 43 49 42 - - Actin, aortic smooth muscle P62737 P62736 0.031 42 kDa 0 1 1 2 2 3 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Meprin A subunit alpha P28825 Q16819 0.032 84 kDa 17 17 17 19 25 25 [89] - Leucine-rich HEV glycoprotein Q91XL1 P02750 0.034 ? 15 13 11 8 9 9 - - Pro-epidermal growth factor P01132 P01133 0.036 133 kDa 55 56 59 68 72 62 - - Fibronectin P11276 P02751 0.038 273 kDa 3 2 0 6 7 4 [34] [33] Kng2 protein Q6S9I0 no 0.039 47 kDa 30 30 30 42 34 36 - - Alpha-1-acid glycoprotein 1 Q60590 P02763 0.039 24 kDa 4 3 2 6 5 5 - - MCG15829, isoform CRA_a Q3KQQ2 no 0.04 ? 8 11 9 5 4 7 - - Ceruloplasmin Q61147 P00450 0.044 121 kDa 0 1 0 2 2 4 [31, 32] [29] N-acetylglucosamine-6-sulfatase Q8BFR4 P15586 0.048 61 kDa 3 6 3 1 1 2 - - Mammalian ependymin-related protein 1 Q99M71 Q9UM22 0.0001 25 kDa 2 2 2 0 0 0 - - N-acyl-aromatic-L-amino acid amidohydrolase 0.0001 - - Q91XE4 Q96HD9 35 kDa 7 6 6 0 0 0 (carboxylate-forming) 7 0.0002 - - Pancreatic alpha-amylase P00688 no 57 kDa 7 2 2 28 29 27 2 0.0002 - - Protein Ighg2c F6TQW2 no ? 3 4 5 10 11 10 4 0.0002 - - Alpha-2-macroglobulin Q61838 no 166 kDa 21 21 17 1 1 1 6 Glycosylation-dependent cell adhesion molecule 1 Q02596 no 0.0005 16 kDa 0 0 0 2 1 2 - - 30-50 kDa 0.0007 [99-101] - Prothrombin P19221 P00734 70 kDa 4 4 5 7 8 7 3 Cluster of Lysosomal thioesterase PPT2 O35448 Q9UMR5 0.0015 34 kDa 3 2 3 0 0 0 - - Urokinase-type plasminogen activator P06869 P00749 0.0023 48 kDa 6 8 7 17 13 16 [64] - Nidogen-2 O88322 Q14112 0.0027 154 kDa 0 0 0 2 1 1 - - Prostatic spermine-binding protein P15501 no 0.0034 22 kDa 3 3 5 0 0 0 - - dorant-binding protein 2a Q8K1H9 no 0.0052 20 kDa 12 7 8 1 1 1 - - Deoxyribonuclease-1 P49183 P24855 0.0053 32 kDa 22 24 24 14 17 12 - - Complement factor D P03953 P00746 0.0062 28 kDa 20 41 28 78 78 62 - - UPF0762 protein C6orf58 homolog Q8C6C9 Q6P5S2 0.0095 38 kDa 2 2 1 0 0 0 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Napsin-A O09043 O96009 0.0099 46 kDa 12 11 10 14 13 14 [97] Major urinary protein 20 Q5FW60 no 0.013 21 kDa 16 17 20 9 11 6 - - L-lactate dehydrogenase B chain P16125 P07195 0.016 37 kDa 1 2 1 0 0 0 - - Procollagen C-endopeptidase enhancer 1 Q61398 Q15113 0.016 50 kDa 1 2 2 8 9 5 - - Leucine-rich HEV glycoprotein Q91XL1 P02750 0.017 ? 6 5 7 4 3 2 - - Hemoglobin subunit alpha P01942 P69905 0.019 15 kDa 2 5 3 0 0 0 - - Hemoglobin subunit beta-1 P02088 P68871 0.022 16 kDa 2 6 6 0 0 0 - - Lysosomal protective protein P16675 P10619 0.022 54 kDa 1 3 3 0 0 0 [102] - Uromodulin Q91X17 P07911 0.022 71 kDa 11 8 6 21 17 30 - - Prostate stem cell antigen P57096 O43653 0.028 13 kDa 0 0 1 2 1 3 [103] - Cathepsin B P10605 P07858 0.03 37 kDa 3 1 2 0 0 0 [35] [13] Meprin A subunit alpha P28825 Q16819 0.03 84 kDa 7 7 11 14 12 16 [89] - Insulin-like growth factor-binding protein 7 Q61581 Q16270 0.032 29 kDa 5 5 5 6 7 8 [104] - MCG140531 A2AV72 no 0.033 ? 10 8 8 5 6 6 - - Ig kappa chain V19-17 P01633 P06312 0.038 16 kDa 1 1 1 1 0 0 - - Cathepsin Z Q9WUU7 Q9UBR2 0.044 34 kDa 4 2 5 0 0 2 [105] - Odorant-binding protein 2a Q8K1H9 no 0.004 20 kDa 2 2 1 0 0 0 - - Lymphocyte antigen 6D P35459 Q14210 0.0048 13 kDa 1 0 0 3 2 2 - - Napsin-A O09043 O96009 0.0068 46 kDa 1 1 0 3 3 2 [97] - Complement factor D P03953 P00746 0.0078 28 kDa 1 2 3 8 14 11 - - 20-30 kDa Kallikrein-1 P15947 P06870 0.008 29 kDa 0 0 0 5 7 10 [25] [26] Serum albumin P07724 P02768 0.01 69 kDa 2 1 3 5 5 7 [57] [56] 14 - - Cluster of MCG15829, isoform CRA_a Q3KQQ2 no 0.018 ? 151 121 107 88 94 6 Pro-epidermal growth factor P01132 P01133 0.033 133 kDa 1 0 0 5 7 2 - - Major urinary protein 20 Q5FW60 no 0.0012 21 kDa 20 15 17 5 2 5 - - Sulfated glycoprotein 1 Q61207 P07602 0.003 61 kDa 16 16 13 8 6 6 - [40] <20 kDa Clusterin Q06890 P10909 0.0034 52 kDa 0 1 0 3 3 4 [58, 106] [107] Lysozyme C-2 P08905 P61626 0.0049 17 kDa 0 0 0 4 3 2 - - bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Uromodulin Q91X17 P07911 0.0083 71 kDa 1 2 1 5 7 7 - - Complement factor D P03953 P00746 0.011 28 kDa 6 7 6 16 23 28 - - Cystatin-C P21460 P01034 0.011 16 kDa 0 0 0 2 1 1 - [62, 108] Ubiquitin-60S ribosomal protein L40 P62984 P62987 0.011 15 kDa 0 0 0 2 1 1 - - Cluster of MCG116526 D3YYY1 no 0.029 ? 10 6 5 2 2 1 - - Fibronectin P11276 P02751 0.038 273 kDa 0 1 1 4 2 5 [34] [33] Protein Col6a3 E9PWQ3 P12111 0.038 ? 0 0 0 3 1 1 - - Pancreatic alpha-amylase P00688 no 0.046 57 kDa 0 1 1 5 2 3 - - Plasminogen P20918 P00747 0.047 91 kDa 0 2 1 5 5 3 [109] [110]

bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

References

1. Chen-Chen Tana J-TY, and Lan Tan. Biomarkers for Preclinical Alzheimer's Disease. Journal of Alzheimer’s Disease. 2014. doi: 10.3233/JAD-140843. 2. Gao Y. Opinion: Are Urinary Biomarkers from Clinical Studies Biomarkers of Disease or Biomarkers of Medicine? MOJ Proteomics & Bioinformatics. 2014;1(5):00028. doi: 10.15406/mojpb.2014.01.00028. 3. Gao Y. Urine-an untapped goldmine for biomarker discovery? Science China Life sciences. 2013;56(12):1145-6. doi: 10.1007/s11427-013-4574-1. PubMed PMID: 24271956. 4. Wu J, Guo Z, Gao Y. Dynamic changes of urine proteome in a Walker 256 tumor-bearing rat model. Cancer Med. 2017. doi: 10.1002/cam4.1225. PubMed PMID: 28980450. 5. Zhao M, Li M, Li X, Shao C, Yin J, Gao Y. Dynamic changes of urinary proteins in a focal segmental glomerulosclerosis rat model. Proteome science. 2014;12:42. doi: 10.1186/1477-5956-12-42. PubMed PMID: 25061428; PubMed Central PMCID: PMC4109389. 6. Shevchenko G, Wetterhall M, Bergquist J, Hoglund K, Andersson LI, Kultima K. Longitudinal characterization of the brain proteomes for the tg2576 amyloid mouse model using shotgun based mass spectrometry. J Proteome Res. 2012;11(12):6159-74. doi: 10.1021/pr300808h. PubMed PMID: 23050487. 7. Peng J, Guo K, Xia J, Zhou J, Yang J, Westaway D, et al. Development of isotope labeling liquid chromatography mass spectrometry for mouse urine metabolomics: quantitative metabolomic study of transgenic mice related to Alzheimer's disease. J Proteome Res. 2014;13(10):4457-69. doi: 10.1021/pr500828v. PubMed PMID: 25164377. 8. Hu ZP, Browne ER, Liu T, Angel TE, Ho PC, Chan EC. Metabonomic profiling of TASTPM transgenic Alzheimer's disease mouse model. J Proteome Res. 2012;11(12):5903-13. doi: 10.1021/pr300666p. PubMed PMID: 23078235. 9. Young Chul Youn M, PhD, Kun-Woo Park, MD, PhD, Seol-Heui Han, MD, PhD, and SangYun Kim, MD, PhD. Urine neural thread protein measurements in Alzheimer disease. J Am Med Dir Assoc 2011;12:372–6. doi: 10.1016/j.jamda.2010.03.004. PubMed PMID: 21450171. 10. Gao YNaY. Should We Search for Early Brain Disease Biomarkers in Urine.pdf. Authors Journal. 2016;1(1):1-9. doi: 10.15406/aj.2016.01.00003. 11. An M, Gao Y. Urinary Biomarkers of Brain Diseases. Genomics, proteomics & bioinformatics. 2015;13(6):345-54. doi: 10.1016/j.gpb.2015.08.005. PubMed PMID: 26751805; PubMed Central PMCID: PMC4747650. 12. Ni Y, Zhang F, An M, Yin W, Gao Y. Early candidate biomarkers found from urine of astrocytoma rat before changes in MRI. 2017. doi: 10.1101/117333. 13. Sun Y, Rong X, Lu W, Peng Y, Li J, Xu S, et al. Translational study of Alzheimer's disease (AD) biomarkers from brain tissues in AbetaPP/PS1 mice and serum of AD patients. J Alzheimers Dis. 2015;45(1):269-82. doi: 10.3233/JAD-142805. PubMed PMID: 25502766. 14. Laursen B, Mork A, Plath N, Kristiansen U, Bastlund JF. Cholinergic degeneration is associated with increased plaque deposition and cognitive impairment in APPswe/PS1dE9 mice. Behav Brain Res. 2013;240:146-52. doi: 10.1016/j.bbr.2012.11.012. PubMed PMID: 23178660. 15. Kilgore M, Miller CA, Fass DM, Hennig KM, Haggarty SJ, Sweatt JD, et al. Inhibitors of class 1 bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

histone deacetylases reverse contextual memory deficits in a mouse model of Alzheimer's disease. Neuropsychopharmacology. 2010;35(4):870-80. doi: 10.1038/npp.2009.197. PubMed PMID: 20010553; PubMed Central PMCID: PMCPMC3055373. 16. Holcomb L, Gordon MN, McGowan E, Yu X, Benkovic S, Jantzen P, et al. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med. 1998;4(1):97-100. Epub 1998/01/14. PubMed PMID: 9427614. 17. Garcia-Alloza M, Robbins EM, Zhang-Nunes SX, Purcell SM, Betensky RA, Raju S, et al. Characterization of amyloid deposition in the APPswe/PS1dE9 mouse model of Alzheimer disease. Neurobiol Dis. 2006;24(3):516-24. doi: 10.1016/j.nbd.2006.08.017. PubMed PMID: 17029828. 18. Janus C, Flores AY, Xu G, Borchelt DR. Behavioral abnormalities in APPSwe/PS1dE9 mouse model of AD-like pathology: comparative analysis across multiple behavioral domains. Neurobiol Aging. 2015;36(9):2519-32. doi: 10.1016/j.neurobiolaging.2015.05.010. PubMed PMID: 26089165. 19. Zhang W, Hao J, Liu R, Zhang Z, Lei G, Su C, et al. Soluble Abeta levels correlate with cognitive deficits in the 12-month-old APPswe/PS1dE9 mouse model of Alzheimer's disease. Behav Brain Res. 2011;222(2):342-50. doi: 10.1016/j.bbr.2011.03.072. PubMed PMID: 21513747. 20. Bonardi C, de Pulford F, Jennings D, Pardon MC. A detailed analysis of the early context extinction deficits seen in APPswe/PS1dE9 female mice and their relevance to preclinical Alzheimer's disease. Behav Brain Res. 2011;222(1):89-97. doi: 10.1016/j.bbr.2011.03.041. PubMed PMID: 21440575. 21. Speicher KD, Kolbas O, Harper S, Speicher DW. Systematic analysis of peptide recoveries from in-gel digestions for protein identifications in proteome studies. J Biomol Tech. 2000;11(2):74-86. Epub 2000/06/01. PubMed PMID: 19499040; PubMed Central PMCID: PMCPMC2291619. 22. Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nat Methods. 2009;6(5):359-62. Epub 2009/04/21. doi: 10.1038/nmeth.1322. PubMed PMID: 19377485. 23. Twigt JM, Bezstarosti K, Demmers J, Lindemans J, Laven JS, Steegers-Theunissen RP. Preconception folic acid use influences the follicle fluid proteome. European journal of clinical investigation. 2015;45(8):833-41. doi: 10.1111/eci.12478. PubMed PMID: 26094490. 24. Stoop MP, Singh V, Stingl C, Martin R, Khademi M, Olsson T, et al. Effects of natalizumab treatment on the cerebrospinal fluid proteome of multiple sclerosis patients. J Proteome Res. 2013;12(3):1101-7. Epub 2013/01/24. doi: 10.1021/pr3012107. PubMed PMID: 23339689. 25. Buck. TAVaHS. Kallikrein-Kinin System Mediated Inflammation in Alzheimer's Disease In Vivo. . Current Alzheimer Research. 2011;8:59-66. PubMed PMID: 21143155. 26. Diamandis EP, Yousef GM, Petraki C, Soosaipillai AR. Human kallikrein 6 as a biomarker of alzheimer's disease. Clin Biochem. 2000;33(8):663-7. Epub 2001/02/13. PubMed PMID: 11166014. 27. Marshall MS, Bongarzone ER. Beyond Krabbe's disease: The potential contribution of galactosylceramidase deficiency to neuronal vulnerability in late-onset synucleinopathies. J Neurosci Res. 2016;94(11):1328-32. Epub 2016/09/18. doi: 10.1002/jnr.23751. PubMed PMID: 27638614; PubMed Central PMCID: PMCPMC5027968. 28. Filippov V, Song MA, Zhang K, Vinters HV, Tung S, Kirsch WM, et al. Increased ceramide in brains with Alzheimer's and other neurodegenerative diseases. J Alzheimers Dis. 2012;29(3):537-47. Epub 2012/01/20. doi: 10.3233/JAD-2011-111202. PubMed PMID: 22258513; PubMed Central bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

PMCID: PMCPMC3643694. 29. Park JH, Lee DW, Park KS. Elevated serum copper and ceruloplasmin levels in Alzheimer's disease. Asia Pac Psychiatry. 2014;6(1):38-45. Epub 2013/07/17. doi: 10.1111/appy.12077. PubMed PMID: 23857910. 30. Kessler H, Pajonk FG, Meisser P, Schneider-Axmann T, Hoffmann KH, Supprian T, et al. Cerebrospinal fluid diagnostic markers correlate with lower plasma copper and ceruloplasmin in patients with Alzheimer's disease. J Neural Transm (Vienna). 2006;113(11):1763-9. doi: 10.1007/s00702-006-0485-7. PubMed PMID: 16736242. 31. Siotto M, Simonelli I, Pasqualetti P, Mariani S, Caprara D, Bucossi S, et al. Association Between Serum Ceruloplasmin Specific Activity and Risk of Alzheimer's Disease. J Alzheimers Dis. 2016;50(4):1181-9. Epub 2016/02/03. doi: 10.3233/JAD-150611. PubMed PMID: 26836154. 32. Bush AI. The metal theory of Alzheimer's disease. J Alzheimers Dis. 2013;33 Suppl 1:S277-81. Epub 2012/05/29. doi: 10.3233/JAD-2012-129011. PubMed PMID: 22635102. 33. Long J, Pan G, Ifeachor E, Belshaw R, Li X. Discovery of Novel Biomarkers for Alzheimer's Disease from Blood. Dis Markers. 2016;2016:4250480. doi: 10.1155/2016/4250480. PubMed PMID: 27418712; PubMed Central PMCID: PMCPMC4932164. 34. Muenchhoff J, Poljak A, Song F, Raftery M, Brodaty H, Duncan M, et al. Plasma protein profiling of mild cognitive impairment and Alzheimer's disease across two independent cohorts. J Alzheimers Dis. 2015;43(4):1355-73. doi: 10.3233/JAD-141266. PubMed PMID: 25159666. 35. Hook G, Yu J, Toneff T, Kindy M, Hook V. Brain pyroglutamate amyloid-beta is produced by cathepsin B and is reduced by the cysteine protease inhibitor E64d, representing a potential Alzheimer's disease therapeutic. J Alzheimers Dis. 2014;41(1):129-49. doi: 10.3233/JAD-131370. PubMed PMID: 24595198; PubMed Central PMCID: PMCPMC4059604. 36. Mateos L, Ismail MA, Gil-Bea FJ, Leoni V, Winblad B, Bjorkhem I, et al. Upregulation of brain renin angiotensin system by 27-hydroxycholesterol in Alzheimer's disease. J Alzheimers Dis. 2011;24(4):669-79. doi: 10.3233/JAD-2011-101512. PubMed PMID: 21297254. 37. Savaskan E. The Role of the Brain Renin-Angiotensin System in Neurodegenerative Disorders. Current Alzheimer Research. 2005;2:29-35. PubMed PMID: 15977987. 38. McArthur S, Cristante E, Paterno M, Christian H, Roncaroli F, Gillies GE, et al. Annexin A1: a central player in the anti-inflammatory and neuroprotective role of microglia. J Immunol. 2010;185(10):6317-28. doi: 10.4049/jimmunol.1001095. PubMed PMID: 20962261; PubMed Central PMCID: PMCPMC3145124. 39. Yamaguchi M, Kokai Y, Imai S-I, Utsumi K, Matsumoto K, Honda H, et al. Investigation of annexin A5 as a biomarker for Alzheimer's disease using neuronal cell culture and mouse model. Journal of Neuroscience Research. 2010:n/a-n/a. doi: 10.1002/jnr.22427. PubMed PMID: 20648654. 40. Heywood WE, Galimberti D, Bliss E, Sirka E, Paterson RW, Magdalinou NK, et al. Identification of novel CSF biomarkers for neurodegeneration and their validation by a high-throughput multiplexed targeted proteomic assay. Mol Neurodegener. 2015;10:64. doi: 10.1186/s13024-015-0059-y. PubMed PMID: 26627638; PubMed Central PMCID: PMCPMC4666172. 41. Shen L, Chen Y, Yang A, Chen C, Liao L, Li S, et al. Redox Proteomic Profiling of Specifically Carbonylated Proteins in the Serum of Triple Transgenic Alzheimer's Disease Mice. Int J Mol Sci. 2016;17(4):469. doi: 10.3390/ijms17040469. PubMed PMID: 27077851; PubMed Central PMCID: bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

PMCPMC4848925. 42. Zhenwei Shang HL, Mingming Zhang, Lian Duan, Situo Wang, Jin Li, Guiyou Liu, Zhang Ruijie and Yongshuai Jiang. Genome-wide haplotype association study identify TNFRSF1A, CASP7, LRP1B, CDH1 and TG genes associated with Alzheimer’s disease in Caribbean Hispanic individuals. Oncotarget. 2015;6(40):42504-14. PubMed PMID: 26621834. 43. Seong E, Yuan L, Arikkath J. Cadherins and catenins in dendrite and synapse morphogenesis. Cell Adh Migr. 2015;9(3):202-13. doi: 10.4161/19336918.2014.994919. PubMed PMID: 25914083; PubMed Central PMCID: PMCPMC4594442. 44. JOHN J. SR.4MEK NRC, DANIEL J. HURLEY, and RANDALL D. SEIFERT. THE UTILITY OF SALIVARY AMYLASE AS AN EVALUATION OF M MUSCARINIC AGONIST ACTIVITY IN ALSHEIMER’S DISEASE. ProgNeuro-psychopharmacok & Biol Psychiat. 1995;19:85-91. PubMed PMID: 7535938. 45. He X, Huang Y, Li B, Gong CX, Schuchman EH. Deregulation of sphingolipid metabolism in Alzheimer's disease. Neurobiol Aging. 2010;31(3):398-408. doi: 10.1016/j.neurobiolaging.2008.05.010. PubMed PMID: 18547682; PubMed Central PMCID: PMCPMC2829762. 46. Bacchetti T, Vignini A, Giulietti A, Nanetti L, Provinciali L, Luzzi S, et al. Higher Levels of Oxidized Low Density Lipoproteins in Alzheimer's Disease Patients: Roles for Platelet Activating Factor Acetyl Hydrolase and Paraoxonase-1. J Alzheimers Dis. 2015;46(1):179-86. doi: 10.3233/JAD-143096. PubMed PMID: 25720407. 47. Jang BG, Yun SM, Ahn K, Song JH, Jo SA, Kim YY, et al. Plasma carbonic anhydrase II protein is elevated in Alzheimer's disease. J Alzheimers Dis. 2010;21(3):939-45. doi: 10.3233/JAD-2010-100384. PubMed PMID: 20634585. 48. Huang Y, Tanimukai H, Liu F, Iqbal K, Grundke-Iqbal I, Gong CX. Elevation of the level and activity of acid ceramidase in Alzheimer's disease brain. Eur J Neurosci. 2004;20(12):3489-97. doi: 10.1111/j.1460-9568.2004.03852.x. PubMed PMID: 15610181. 49. Gong H, Dong W, Rostad SW, Marcovina SM, Albers JJ, Brunzell JD, et al. Lipoprotein lipase (LPL) is associated with neurite pathology and its levels are markedly reduced in the dentate gyrus of Alzheimer's disease brains. J Histochem Cytochem. 2013;61(12):857-68. doi: 10.1369/0022155413505601. PubMed PMID: 24004859; PubMed Central PMCID: PMCPMC3840745. 50. Zhao Y, Tan W, Sheng W, Li X. Identification of Biomarkers Associated With Alzheimer's Disease by Bioinformatics Analysis. Am J Alzheimers Dis Other Demen. 2016;31(2):163-8. doi: 10.1177/1533317515588181. PubMed PMID: 26082458. 51. Satoh J, Tabunoki H, Ishida T, Saito Y, Arima K. Accumulation of a repulsive axonal guidance molecule RGMa in amyloid plaques: a possible hallmark of regenerative failure in Alzheimer's disease brains. Neuropathol Appl Neurobiol. 2013;39(2):109-20. doi: 10.1111/j.1365-2990.2012.01281.x. PubMed PMID: 22582881. 52. Goedert M, Ghetti B, Spillantini MG. Frontotemporal dementia: implications for understanding Alzheimer disease. Cold Spring Harb Perspect Med. 2012;2(2):a006254. doi: 10.1101/cshperspect.a006254. PubMed PMID: 22355793; PubMed Central PMCID: PMCPMC3281593. 53. Lizard G, Rouaud O, Demarquoy J, Cherkaoui-Malki M, Iuliano L. Potential roles of peroxisomes bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

in Alzheimer's disease and in dementia of the Alzheimer's type. J Alzheimers Dis. 2012;29(2):241-54. doi: 10.3233/JAD-2011-111163. PubMed PMID: 22433776. 54. Silva PN, Furuya TK, Braga IL, Rasmussen LT, Labio RW, Bertolucci PH, et al. Analysis of HSPA8 and HSPA9 mRNA expression and promoter methylation in the brain and blood of Alzheimer's disease patients. J Alzheimers Dis. 2014;38(1):165-70. doi: 10.3233/JAD-130428. PubMed PMID: 23948933. 55. Soilu-Hanninen M, Broberg E, Roytta M, Mattila P, Rinne J, Hukkanen V. Expression of LIF and LIF receptor beta in Alzheimer's and Parkinson's diseases. Acta Neurol Scand. 2010;121(1):44-50. doi: 10.1111/j.1600-0404.2009.01179.x. PubMed PMID: 20074285. 56. Fu Y, Zhao D, Pan B, Wang J, Cui Y, Shi F, et al. Proteomic Analysis of Protein Expression Throughout Disease Progression in a Mouse Model of Alzheimer's Disease. J Alzheimers Dis. 2015;47(4):915-26. doi: 10.3233/JAD-150312. PubMed PMID: 26401771. 57. Yamamoto K, Shimada H, Koh H, Ataka S, Miki T. Serum levels of albumin-amyloid beta complexes are decreased in Alzheimer's disease. Geriatr Gerontol Int. 2014;14(3):716-23. doi: 10.1111/ggi.12147. PubMed PMID: 24020590. 58. Miners JS, Clarke P, Love S. Clusterin levels are increased in Alzheimer's disease and influence the regional distribution of Abeta. Brain Pathol. 2017;27(3):305-13. doi: 10.1111/bpa.12392. PubMed PMID: 27248362. 59. Prikrylova Vranova H, Henykova E, Mares J, Kaiserova M, Mensikova K, Vastik M, et al. Clusterin CSF levels in differential diagnosis of neurodegenerative disorders. J Neurol Sci. 2016;361:117-21. doi: 10.1016/j.jns.2015.12.023. PubMed PMID: 26810527. 60. Hu WT, Watts KD, Tailor P, Nguyen TP, Howell JC, Lee RC, et al. CSF complement 3 and factor H are staging biomarkers in Alzheimer's disease. Acta Neuropathol Commun. 2016;4:14. doi: 10.1186/s40478-016-0277-8. PubMed PMID: 26887322; PubMed Central PMCID: PMCPMC4758165. 61. Lian H, Yang L, Cole A, Sun L, Chiang AC, Fowler SW, et al. NF-kappaB-activated astroglial release of complement C3 compromises neuronal morphology and function associated with Alzheimer's disease. Neuron. 2015;85(1):101-15. doi: 10.1016/j.neuron.2014.11.018. PubMed PMID: 25533482; PubMed Central PMCID: PMCPMC4289109. 62. Perrin RJ, Craig-Schapiro R, Malone JP, Shah AR, Gilmore P, Davis AE, et al. Identification and validation of novel cerebrospinal fluid biomarkers for staging early Alzheimer's disease. PLoS One. 2011;6(1):e16032. doi: 10.1371/journal.pone.0016032. PubMed PMID: 21264269; PubMed Central PMCID: PMCPMC3020224. 63. Uberti D, Cenini G, Bonini SA, Barcikowska M, Styczynska M, Szybinska A, et al. Increased CD44 gene expression in lymphocytes derived from Alzheimer disease patients. Neurodegener Dis. 2010;7(1-3):143-7. doi: 10.1159/000289225. PubMed PMID: 20197694. 64. Wu W, Jiang H, Wang M, Zhang D. Meta-analysis of the association between urokinase-plasminogen activator gene rs2227564 polymorphism and Alzheimer's disease. Am J Alzheimers Dis Other Demen. 2013;28(5):517-23. doi: 10.1177/1533317513494450. PubMed PMID: 23813610. 65. Liu Y, Cao X. Characteristics and Significance of the Pre-metastatic Niche. Cancer Cell. 2016;30(5):668-81. doi: 10.1016/j.ccell.2016.09.011. PubMed PMID: 27846389. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

66. Hamilton LK, Dufresne M, Joppe SE, Petryszyn S, Aumont A, Calon F, et al. Aberrant Lipid Metabolism in the Forebrain Niche Suppresses Adult Neural Stem Cell Proliferation in an Animal Model of Alzheimer's Disease. Cell Stem Cell. 2015;17(4):397-411. doi: 10.1016/j.stem.2015.08.001. PubMed PMID: 26321199. 67. Liu Q, Zhang J. Lipid metabolism in Alzheimer's disease. Neurosci Bull. 2014;30(2):331-45. doi: 10.1007/s12264-013-1410-3. PubMed PMID: 24733655; PubMed Central PMCID: PMCPMC5562656. 68. Giannopoulos PF, Joshi YB, Pratico D. Novel lipid signaling pathways in Alzheimer's disease pathogenesis. Biochem Pharmacol. 2014;88(4):560-4. doi: 10.1016/j.bcp.2013.11.005. PubMed PMID: 24269629; PubMed Central PMCID: PMCPMC3972350. 69. Mario D ́ıaz NıF, Virginia Mart ́ın, Isidre Ferrer, Toma ́s Go ́mez and Raquel Mar ́ın. Biophysical Alterations in Lipid Rafts from Human Cerebral Cortex Associate with Increased BACE1/A PP Interaction in Early Stages of Alzheimer’s Disease. Journal of Alzheimer’s Disease. 2015;43:1185–98. PubMed PMID: 25147112. 70. Tai J, Liu W, Li Y, Li L, Holscher C. Neuroprotective effects of a triple GLP-1/GIP/glucagon receptor agonist in the APP/PS1 transgenic mouse model of Alzheimer's disease. Brain Res. 2017;1678:64-74. doi: 10.1016/j.brainres.2017.10.012. PubMed PMID: 29050859. 71. Wang ZX, Tan L, Yu JT. Axonal transport defects in Alzheimer's disease. Mol Neurobiol. 2015;51(3):1309-21. doi: 10.1007/s12035-014-8810-x. PubMed PMID: 25052480. 72. Qian M, Shen X, Wang H. The Distinct Role of ADAM17 in APP Proteolysis and Microglial Activation Related to Alzheimer's Disease. Cell Mol Neurobiol. 2016;36(4):471-82. doi: 10.1007/s10571-015-0232-4. PubMed PMID: 26119306. 73. Cortes-Canteli M, Paul J, Norris EH, Bronstein R, Ahn HJ, Zamolodchikov D, et al. Fibrinogen and beta-amyloid association alters thrombosis and fibrinolysis: a possible contributing factor to Alzheimer's disease. Neuron. 2010;66(5):695-709. doi: 10.1016/j.neuron.2010.05.014. PubMed PMID: 20547128; PubMed Central PMCID: PMCPMC2895773. 74. Gowrishankar S, Yuan P, Wu Y, Schrag M, Paradise S, Grutzendler J, et al. Massive accumulation of luminal protease-deficient axonal lysosomes at Alzheimer's disease amyloid plaques. Proc Natl Acad Sci U S A. 2015;112(28):E3699-708. doi: 10.1073/pnas.1510329112. PubMed PMID: 26124111; PubMed Central PMCID: PMCPMC4507205. 75. Wilfred A Jefferies KAP, Kaan E Biron, Franz Fenninger, Cheryl G Pfeifer and Dara L Dickstein. Adjusting the compass: new insights into the role of angiogenesis in Alzheimer’s disease. Alzheimer's Research & Therapy. 2013;5(6):64. 76. Vagnucci AH, Li WW. Alzheimer's disease and angiogenesis. The Lancet. 2003;361(9357):605-8. doi: 10.1016/s0140-6736(03)12521-4. 77. Qin W, Jia X, Wang F, Zuo X, Wu L, Zhou A, et al. Elevated plasma angiogenesis factors in Alzheimer's disease. J Alzheimers Dis. 2015;45(1):245-52. doi: 10.3233/JAD-142409. PubMed PMID: 25690662. 78. Bredesen. DE. Metabolic profiling distinguishes three subtypes of Alzheimer’s disease. AGING. 2015;7(7):595-600. 79. Stephan BC, Matthews FE, Ma B, Muniz G, Hunter S, Davis D, et al. Alzheimer and vascular neuropathological changes associated with different cognitive States in a non-demented sample. J bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Alzheimers Dis. 2012;29(2):309-18. doi: 10.3233/JAD-2011-110518. PubMed PMID: 22233761; PubMed Central PMCID: PMCPMC3975483. 80. Howard J PG. Fibronectin staining detects micro-organisms in aged and Alzheimer's disease brain. Neuroreport 1992;3(3):615-8. 81. Kitamura Y, Usami R, Ichihara S, Kida H, Satoh M, Tomimoto H, et al. Plasma protein profiling for potential biomarkers in the early diagnosis of Alzheimer's disease. Neurol Res. 2017;39(3):231-8. doi: 10.1080/01616412.2017.1281195. PubMed PMID: 28107809. 82. Boddaert J, Kinugawa K, Lambert JC, Boukhtouche F, Zoll J, Merval R, et al. Evidence of a role for lactadherin in Alzheimer's disease. Am J Pathol. 2007;170(3):921-9. doi: 10.2353/ajpath.2007.060664. PubMed PMID: 17322377; PubMed Central PMCID: PMCPMC1864868. 83. A.V. Alessenko AEBaLBD. Connection of lipid peroxide oxidation with the sphingomyelin pathway in the development of Alzheimer’s disease. Biochemical Society Transactions. 2004;32(1):144-6. PubMed PMID: 14748735. 84. Rygiel K. Can angiotensin-converting enzyme inhibitors impact cognitive decline in early stages of Alzheimer's disease? An overview of research evidence in the elderly patient population. Journal of Postgraduate Medicine. 2016;62(4):242-8. PubMed PMID: 27763482 85. Gard PR, Fidalgo S, Lotter I, Richardson C, Farina N, Rusted J, et al. Changes of renin-angiotensin system-related aminopeptidases in early stage Alzheimer's disease. Exp Gerontol. 2017;89:1-7. doi: 10.1016/j.exger.2017.01.006. PubMed PMID: 28069385. 86. Kaur G, Pawlik M, Gandy SE, Ehrlich ME, Smiley JF, Levy E. Lysosomal dysfunction in the brain of a mouse model with intraneuronal accumulation of carboxyl terminal fragments of the amyloid precursor protein. Mol Psychiatry. 2017;22(7):981-9. doi: 10.1038/mp.2016.189. PubMed PMID: 27777419; PubMed Central PMCID: PMCPMC5405008. 87. Armstrong A, Mattsson N, Appelqvist H, Janefjord C, Sandin L, Agholme L, et al. Lysosomal network proteins as potential novel CSF biomarkers for Alzheimer's disease. Neuromolecular Med. 2014;16(1):150-60. doi: 10.1007/s12017-013-8269-3. PubMed PMID: 24101586; PubMed Central PMCID: PMCPMC3918123. 88. Harrop EMPMHAR. Cortical glutaminase, beta-glucuronidase and glucose utilization in Alzheimer's disease. Can J Neurol Sci 1989;16(4 Suppl):511-5. 89. Bien J, Jefferson T, Causevic M, Jumpertz T, Munter L, Multhaup G, et al. The metalloprotease meprin beta generates amino terminal-truncated amyloid beta peptide species. J Biol Chem. 2012;287(40):33304-13. doi: 10.1074/jbc.M112.395608. PubMed PMID: 22879596; PubMed Central PMCID: PMCPMC3460434. 90. Morello N, Tonoli E, Logrand F, Fiorito V, Fagoonee S, Turco E, et al. Haemopexin affects iron distribution and ferritin expression in mouse brain. J Cell Mol Med. 2009;13(10):4192-204. doi: 10.1111/j.1582-4934.2008.00611.x. PubMed PMID: 19120692; PubMed Central PMCID: PMCPMC4496126. 91. Manral P SP, Hariprasad G, Chandralekha, Tripathi M, Srinivasan A. Can apolipoproteins and complement factors be biomarkers of Alzheimer's disease? Curr Alzheimer Res. 2012;9(8):935-43. 92. Puertas Mdel C, Martinez-Martos JM, Cobo M, Lorite P, Sandalio RM, Palomeque T, et al. Plasma renin-angiotensin system-regulating aminopeptidase activities are modified in early stage bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Alzheimer's disease and show gender differences but are not related to apolipoprotein E genotype. Exp Gerontol. 2013;48(6):557-64. doi: 10.1016/j.exger.2013.03.002. PubMed PMID: 23500679. 93. Chapuis J, Vingtdeux V, Campagne F, Davies P, Marambaud P. Growth arrest-specific 1 binds to and controls the maturation and processing of the amyloid-beta precursor protein. Hum Mol Genet. 2011;20(10):2026-36. doi: 10.1093/hmg/ddr085. PubMed PMID: 21357679; PubMed Central PMCID: PMCPMC3279048. 94. Hare DJ, Doecke JD, Faux NG, Rembach A, Volitakis I, Fowler CJ, et al. Decreased plasma iron in Alzheimer's disease is due to transferrin desaturation. ACS Chem Neurosci. 2015;6(3):398-402. doi: 10.1021/cn5003557. PubMed PMID: 25588002. 95. Haldar S, Beveridge J, Wong J, Singh A, Galimberti D, Borroni B, et al. A low-molecular-weight ferroxidase is increased in the CSF of sCJD cases: CSF ferroxidase and transferrin as diagnostic biomarkers for sCJD. Antioxid Redox Signal. 2013;19(14):1662-75. doi: 10.1089/ars.2012.5032. PubMed PMID: 23379482; PubMed Central PMCID: PMCPMC3809602. 96. Lieberman J SL, Tachiki KH, Kling AS. Serum alpha 1-antichymotrypsin level as a marker for Alzheimer-type dementia. Neurobiol Aging 1995;16(5):747-53. 97. Gruninger-Leitch F, Schlatter D, Kung E, Nelbock P, Dobeli H. Substrate and inhibitor profile of BACE (beta-secretase) and comparison with other mammalian aspartic proteases. J Biol Chem. 2002;277(7):4687-93. doi: 10.1074/jbc.M109266200. PubMed PMID: 11741910. 98. Daborg J, Andreasson U, Pekna M, Lautner R, Hanse E, Minthon L, et al. Cerebrospinal fluid levels of complement proteins C3, C4 and CR1 in Alzheimer's disease. J Neural Transm (Vienna). 2012;119(7):789-97. doi: 10.1007/s00702-012-0797-8. PubMed PMID: 22488444. 99. Akiyama H IK, Kondo H, McGeer PL. Thrombin accumulation in brains of patients with Alzheimer's disease. Neurosci Lett 1992;146(2):152-4. PubMed PMID: 1491781. 100. Arai T MJ, Klegeris A, Guo JP, McGeer PL. Thrombin and prothrombin are expressed by neurons and glial cells and accumulate in neurofibrillary tangles in Alzheimer disease brain. J Neuropathol Exp Neurol 2006;65(1):19-25. PubMed PMID: 16410745. 101. Lewczuk P WJ, Lange M, Jahn H, Reiber H, Ehrenreich H. Prothrombin concentration in the cerebrospinal fluid is not altered in Alzheimer's disease. Neurochem Res 1999;24(12):1531-4. PubMed PMID: 10591402. 102. Butler D, Hwang J, Estick C, Nishiyama A, Kumar SS, Baveghems C, et al. Protective effects of positive lysosomal modulation in Alzheimer's disease transgenic mouse models. PLoS One. 2011;6(6):e20501. doi: 10.1371/journal.pone.0020501. PubMed PMID: 21695208; PubMed Central PMCID: PMCPMC3112200. 103. Jensen MM, Arvaniti M, Mikkelsen JD, Michalski D, Pinborg LH, Hartig W, et al. Prostate stem cell antigen interacts with nicotinic acetylcholine receptors and is affected in Alzheimer's disease. Neurobiol Aging. 2015;36(4):1629-38. doi: 10.1016/j.neurobiolaging.2015.01.001. PubMed PMID: 25680266. 104. Agbemenyah HY, Agis-Balboa RC, Burkhardt S, Delalle I, Fischer A. Insulin growth factor binding protein 7 is a novel target to treat dementia. Neurobiol Dis. 2014;62:135-43. doi: 10.1016/j.nbd.2013.09.011. PubMed PMID: 24075854. 105. Hafner A, Glavan G, Obermajer N, Zivin M, Schliebs R, Kos J. Neuroprotective role of gamma-enolase in microglia in a mouse model of Alzheimer's disease is regulated by cathepsin X. bioRxiv preprint doi: https://doi.org/10.1101/258921; this version posted February 2, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Aging Cell. 2013;12(4):604-14. doi: 10.1111/acel.12093. PubMed PMID: 23621429. 106. Vishnu VY, Modi M, Sharma S, Mohanty M, Goyal MK, Lal V, et al. Role of Plasma Clusterin in Alzheimer's Disease-A Pilot Study in a Tertiary Hospital in Northern India. PLoS One. 2016;11(11):e0166369. doi: 10.1371/journal.pone.0166369. PubMed PMID: 27861589; PubMed Central PMCID: PMCPMC5115728. 107. Liang HC, Russell C, Mitra V, Chung R, Hye A, Bazenet C, et al. Glycosylation of Human Plasma Clusterin Yields a Novel Candidate Biomarker of Alzheimer's Disease. J Proteome Res. 2015;14(12):5063-76. doi: 10.1021/acs.jproteome.5b00892. PubMed PMID: 26488311. 108. Choi YS, Hou S, Choe LH, Lee KH. Targeted human cerebrospinal fluid proteomics for the validation of multiple Alzheimer's disease biomarker candidates. J Chromatogr B Analyt Technol Biomed Life Sci. 2013;930:129-35. doi: 10.1016/j.jchromb.2013.05.003. PubMed PMID: 23735279; PubMed Central PMCID: PMCPMC3710693. 109. Hanzel CE, Iulita MF, Eyjolfsdottir H, Hjorth E, Schultzberg M, Eriksdotter M, et al. Analysis of matrix metallo-proteases and the plasminogen system in mild cognitive impairment and Alzheimer's disease cerebrospinal fluid. J Alzheimers Dis. 2014;40(3):667-78. doi: 10.3233/JAD-132282. PubMed PMID: 24531161. 110. Oh J, Lee HJ, Song JH, Park SI, Kim H. Plasminogen activator inhibitor-1 as an early potential diagnostic marker for Alzheimer's disease. Exp Gerontol. 2014;60:87-91. doi: 10.1016/j.exger.2014.10.004. PubMed PMID: 25304332.